Modelling Cellular Networks for Global Spectrum Management

International Spectrum Management at the International Telecommunication Union (ITU) is a complex puzzle. A critical piece of this puzzle is ensuring the coexistence of diverse radiocommunication services, from satellite systems to scientific radars, with the ever-expanding footprint of terrestrial International Mobile Telecommunication (IMT) networks, commonly referred to as cellular networks. The central challenge we face is twofold, technical and inherently philosophical:

 How do we model these complex cellular networks with accuracy and realism, while maintaining the genericity required for global regulations and policymaking?

While mobile network operators focus on hyper-accurate, site-specific planning to guarantee Quality of Service (QoS) to their subscribers, the ITU’s mission is broader: to allocate spectrum fairly and efficiently on a worldwide basis. This requires a robust, standardized, and repeatable methodology.

The ITU-R has developed a suite of Recommendations and Reports that provide this framework. They offer guidance on generically modelling IMT across diverse environments, from dense urban to rural. This includes defining base stations densities, heights, user distributions to simulate network load, and algorithms to encapsulate the latest technological features like Active Antenna Systems (AAS) and power control algorithms.

But how is this abstract and sometimes scattered guidance translated into a practical analysis? Let us break down the process.

Step 1: Network Deployment, Art of the Possible

The first step involves deploying a representative cellular network within a specific geographic area. The area itself is not the focus. Rather, it serves as a realistic and representative environment for modelling the network's behaviour under typical conditions.

The image below shows a model of an urban deployment in Shibaura, Minato City, Tokyo. Cells are represented with different colours. A key distinction from commercial planning is the assumption of a regular cell grid. Real-world deployments meticulously avoid shadowing from buildings and other obstacles, but in the context of global spectrum management, we prioritize a generic and transferrable approach over site-specific optimisation. This ensures our conclusions are based on the typical network behaviour, not the specificities of a single location.

Simulated IMT deployment mapped onto a real-work urban environment

Step 2: Base Stations Locations and Choice of Antennas, the Engine Room

The most prevalent model uses tri-sectorized base stations situated at the intersection of three cells. This model maximizes sites re-use and is far more common than omnidirectional deployments, where a single base station is located at the centre of each cell.

The choice of antenna model is crucial. We must decide between traditional static antennas and modern AAS. AAS, as illustrated in the figure below, uses an array of individually phased radiating elements to form dynamic, user-specific beams. Modelling this beam-steering capability is essential for accurate 5G-and-beyond analysis as it dramatically focuses energy and reduces inter-cell interference effects.

Step 3: Modelling the Link: A Two-Way Street

Modelling connections between users and base stations is more than just plotting a line on a map. It requires:

• Advanced Propagation Models: Using terrain data to estimate real-world path loss, factoring in diffraction, reflection, and clutter. Recommendation ITU-R P.452-18 provides a systematic methodology and is therefore broadly used in this context.

•  Closed-loop power control: In the uplink, user devices constantly adapt their power to maintain the signal received at the connected base station at a constant level, saving battery and minimising interference to neighbouring cells. As shown in the image below, the result is a vast dynamic range, from a maximum of +23 dBm (200 mW) for cell-edge and indoor users down to a mere –30 dBm (1 μW) for those with excellent signal conditions.

Step 4: Inter-cell interference, the Invisible Enemy

Perhaps the most critical factor in modelling cellular networks is inter-cell interference. This is the “noise” from users in neighbouring cells that limits capacity and quality. Accurate modelling is vital, as external systems (e.g., a new satellite service) will only add to this interference noise floor.

As shown below; in some cases, the inter-cell interference can even exceed the user’s desired signal, leading to dropped connections or unstable handovers. A realistic model must capture these complex, dynamic interactions to truly understand the network’s resilience to external interference.

Conclusion

Modelling IMT networks in the context of global Spectrum Management requires a balance between realistic complexity and necessary generalisation. It requires powerful and computing-efficient tools to execute correctly.

The good news for you? You won’t have to build this capability from scratch.

At River Advisers, our modelling tools are designed specifically for this task. They precisely implement the ITU’s framework, allowing us to produce repeatable and evidence-based studies in a timely manner, in particular for WRC-27 preparation and other upcoming spectrum management challenges.